3. C O N S T R U C T I O N A N D USE OF S Y N T H E T I C M E D I A
111
The composition of the salt mixture may be that of one of the now
classical formulae (Gey and Gey, 1936; Earle, 1943; Hanks, 1948; Table
II, p. 131) or may be devised for the particular cells under study. Bicarbonate has generally been thought to be essential (Harris, 1954; Swim and
Parker, 1958c; Geyer and Chang, 1958), at least in small amounts, but
is often supplied at high concentrations for buffering purposes. The
probability is high that all cells do need a source of C O a , but that
quite low concentrations may suffice for some cells, especially if it is being
generated in the system. For example, bicarbonate has been omitted
completely in media which include oxaloacetate (Gwatkin and Siminovitch, 1960; Kelley, Adamson and Vail, 1960). Several attempts have
been made to replace bicarbonate, at least in large part, by other buffering
systems, because of the technical complications of dependence of a bicarbonate-containing mixture upon a controlled gas-phase. Swim and Parker
(1955) and Swim (1961) have successfully used tris(hydroxymethyl)aminomethane (Tris; Tham), or phosphite, for several strains of human
and mouse cells.
Varying concentrations of inorganic phosphate have been used,
as well as organic sources of phosphate, in synthetic media. In the
absence of the large amounts of organic phosphates which biological
media supply, inorganic phosphate becomes one of the limiting factors
controlling glycolysis (Waymouth, 1954b; Racker and W u , 1959).
A high phosphate content may therefore be beneficial both for buffering
and for metabolic reasons. Variations in Ca and M g can greatly affect
cell morphology (Owens, Gey and Gey, 1958) and the adhesion of cells
to glass and other substrates. A balance has to be carefully achieved,
between a concentration of Ca sufficient for maintaining the surface
properties of the cell, and one which will, especially in the presence
of a high concentration of phosphate, result in precipitation. The concentration of K may profoundly affect growth rate (Owens, Gey and
Gey, 1956).
The requirements of cells for ions other than the major ones have
been little examined, mainly because this question could not be very
profitably evaluated in media composed of, and supplemented with,
biological materials carrying variable and unknown complements of
trace elements. Now that synthetic media have been developed to the
stage where they can be used without biological supplements, the
pioneer work of Shooter and Gey (1952), which demonstrated the
needs of rat fibroblasts, not only for the major cations Na, K,Ca and
Mg, but also for Fe, Mn, Cu, Co, Zn and M o , has been confirmed and
supported, for human and mouse cells by Melnick, Hsiung, Rappaport,
Howes and Reissig (1957) and by Waymouth (1960a). The individual
ions have not yet all been carefully tested in completely defined media,
111
The composition of the salt mixture may be that of one of the now
classical formulae (Gey and Gey, 1936; Earle, 1943; Hanks, 1948; Table
II, p. 131) or may be devised for the particular cells under study. Bicarbonate has generally been thought to be essential (Harris, 1954; Swim and
Parker, 1958c; Geyer and Chang, 1958), at least in small amounts, but
is often supplied at high concentrations for buffering purposes. The
probability is high that all cells do need a source of C O a , but that
quite low concentrations may suffice for some cells, especially if it is being
generated in the system. For example, bicarbonate has been omitted
completely in media which include oxaloacetate (Gwatkin and Siminovitch, 1960; Kelley, Adamson and Vail, 1960). Several attempts have
been made to replace bicarbonate, at least in large part, by other buffering
systems, because of the technical complications of dependence of a bicarbonate-containing mixture upon a controlled gas-phase. Swim and Parker
(1955) and Swim (1961) have successfully used tris(hydroxymethyl)aminomethane (Tris; Tham), or phosphite, for several strains of human
and mouse cells.
Varying concentrations of inorganic phosphate have been used,
as well as organic sources of phosphate, in synthetic media. In the
absence of the large amounts of organic phosphates which biological
media supply, inorganic phosphate becomes one of the limiting factors
controlling glycolysis (Waymouth, 1954b; Racker and W u , 1959).
A high phosphate content may therefore be beneficial both for buffering
and for metabolic reasons. Variations in Ca and M g can greatly affect
cell morphology (Owens, Gey and Gey, 1958) and the adhesion of cells
to glass and other substrates. A balance has to be carefully achieved,
between a concentration of Ca sufficient for maintaining the surface
properties of the cell, and one which will, especially in the presence
of a high concentration of phosphate, result in precipitation. The concentration of K may profoundly affect growth rate (Owens, Gey and
Gey, 1956).
The requirements of cells for ions other than the major ones have
been little examined, mainly because this question could not be very
profitably evaluated in media composed of, and supplemented with,
biological materials carrying variable and unknown complements of
trace elements. Now that synthetic media have been developed to the
stage where they can be used without biological supplements, the
pioneer work of Shooter and Gey (1952), which demonstrated the
needs of rat fibroblasts, not only for the major cations Na, K,Ca and
Mg, but also for Fe, Mn, Cu, Co, Zn and M o , has been confirmed and
supported, for human and mouse cells by Melnick, Hsiung, Rappaport,
Howes and Reissig (1957) and by Waymouth (1960a). The individual
ions have not yet all been carefully tested in completely defined media,